Related Experiment Video
Updated: Nov 10, 2025
06:53
Detection and Monitoring of Tumor Associated Circulating DNA in Patient Biofluids
Published on: June 8, 2019
9.0K
An AI-Powered Blood Test to Detect Cancer Using NanoDSF.
Philipp O Tsvetkov1,2, Rémi Eyraud3, Stéphane Ayache4
1Faculté des Sciences Médicales et Paramédicales, Inst Neurophysiopathol, CNRS, INP, Aix Marseille Univ, 13005 Marseille, France.
Cancers
|April 3, 2021
Summary
Diagnosing aggressive brain tumors like glioblastoma may soon be possible with a simple blood test. This novel method uses plasma denaturation profiles and machine learning to detect cancer with 92% accuracy.
Area of Science:
- Oncology
- Biochemistry
- Medical Diagnostics
Background:
- Glioblastoma is a highly aggressive primary brain tumor, often necessitating invasive biopsies for diagnosis and monitoring.
- Current diagnostic and monitoring methods for glioblastoma are limited by risks associated with biopsies, especially in deep-seated tumors or patients with comorbidities.
Purpose of the Study:
- To develop a minimally invasive diagnostic and monitoring tool for glioblastoma.
- To establish a novel cancer detection method utilizing plasma denaturation profiles.
Main Methods:
- A non-conventional application of differential scanning fluorimetry was employed to analyze plasma denaturation profiles.
- Machine learning algorithms were utilized to distinguish denaturation profiles between glioma patients and healthy controls.
- Blood samples from 84 glioma patients and 63 healthy controls were analyzed.
Main Results:
- The developed method achieved 92% accuracy in distinguishing plasma denaturation profiles of glioma patients from healthy controls.
- The workflow demonstrated high throughput capability.
- The findings suggest a potential for a powerful pan-cancer diagnostic and monitoring tool.
Conclusions:
- Plasma denaturation profiling via differential scanning fluorimetry offers a promising minimally invasive approach for glioblastoma detection.
- This method, enhanced by machine learning, could serve as a valuable pan-cancer diagnostic and monitoring tool.
- The proposed workflow requires only a simple blood test, reducing the need for invasive procedures.
